Downloads · 30 days
9
24% of all-time downloads
easyminnn/atq-labelgated-2.5x-vlm_contact-60k
atq-labelgated-2.5x-vlm_contact-60k is a robotics model from easyminnn. Use it for the robotics task on the model card, and read the license before you ship it in a product. The card lists the license as other.
Inference-only checkpoint (weights + config; optimizer/scheduler/RNG state stripped). Fine-tuned from nvidia/GR00T-N1.5-3B on RoboCasa MG 300 (7,200 episodes, 3 cameras 256x256, 20 fps) with VLM speed-ratio labels bak…
Downloads · 30 days
9
24% of all-time downloads
All-time downloads
38
Public
Parameters
2.8B
8 GB on disk
Likes
0
Public
Click a slice to open those files.
.safetensors8 GB · 100%
How the weights are stored.
BF161.7B · 58%
From the Hugging Face model README
vlm_contact, 60k stepsInference-only checkpoint (weights + config; optimizer/scheduler/RNG state stripped).
Fine-tuned from nvidia/GR00T-N1.5-3B on RoboCasa MG 300 (7,200 episodes, 3 cameras 256x256, 20 fps)
with VLM speed-ratio labels baked into the action tensor (ratio_label, dims 12:14).
| Experts (4) | main h16 1x (16 rows) · n8 h8 1x (8 rows) · m8 h16 2.5x ([2,3,2,3,2,3] -> 6 rows / 15 steps) · m4 h8 2.5x ([2,3,3] -> 3 rows / 8 steps) |
| Gating | label-gated: conf head decides fine vs compressed group (pred >= tau), router picks horizon inside the group. Gate applies at inference only; all four experts are trained every step |
| conf target | conf * (1 - fixed) (VLM speed-ratio confidence, forced to 0 on contact frames) |
| conf head | scalar regression, MSE, conf_loss_coef=0.1, readout detached from state encoder |
| Router losses | soft_mixture + kl_supervise (0.1) + balance (0.05), router warmup 5,000 steps, min_prob 0.05 |
| Training | 60,000 steps, global batch 64 (32 x 2 GPU), AdamW lr 1e-4 cosine, warmup 0.05, seed 42, bf16, 2x H200 |
| Labels | prehj/robocasa-ratio-labels-contact |
| Code | rakybond007/GR00T-action-quantization branch jimin-dev-label-gated @ 3345064 (scripts/gr00t_finetune.py --use-moe-routing --moe-num-experts 4 --use-merged-8-head --use-merged-4-head --use-native-8-head --moe-label-gated --moe-speed 2.5 --data-config single_panda_gripper_conf) |
tau (conf_threshold) is an eval-time knob; the stored 0.5 is a default, not a trained constant.
Output metadata carries _flevel_level / _flevel_k; the client executes the returned rows as-is (no re-merging).